it is the solution of facial-expression-challenge which is going on kaggle website: link is below ==> (https://www.kaggle.com/c/challenges-in-representation-learning-facial-expression-recognition-challenge) ->install all the depedencies with given version in your PC: dependencies =={ keras==2.2.4 tensorflow==2.0.0-alpha0 opencv==4.1.0 Python==3.7.3 } Download fer2013.csv file from the given link below== https://www.kaggle.com/c/challenges-in-representation-learning-facial-expression-recognition-challenge/data and put fer2013.csv in your root directory +++++++++++ I have already trained this model and store model weights in face_model.h5 and model structure stored in face_model.json.So there is no need to train model again. if you want to train model again then run all the cells of facial-expression-challenge.ipynb. Download fer2013.csv file from the given link below== https://www.kaggle.com/c/challenges-in-representation-learning-facial-expression-recognition-challenge/data and put fer2013.csv in your root directory. ############################ test your model : run camera.py file as: python camera.py